A nonintrusive fast residential load identification algorithm based on frequency‐domain template filtering

Abstract

User‐side load monitoring is a key technology to realize smart utilization of electric power. Since the traditional intrusive load monitoring involves a comparatively large economic cost and execution complexity, this paper studies a way to rapidly identify residential power load in a nonintrusive monitoring mode. A template‐filtering‐based nonintrusive, rapid residential load identification algorithm is proposed, which is based on frequency‐domain analysis of current signals, in combination with the current model when the nonintrusive monitoring load is in operation, and by making use of the fact that the spectrum components of the current signals working independently are completely contained in the hybrid current spectrum. Characteristic currents of the various loads in the power grid are acquired a priori to establish the characteristic filter, and 0–1 valuation is performed on their spectrum components to get the template filter. The template filter is then used to filter the hybrid current signals captured in the nonintrusive mode, and the operation status of the loads is judged and determined after quantification of the filtered frequency components. For the same type of load under different operating conditions, the template filter can be commonly used. Efficiency of the algorithm is verified by making use of the actually collected power consumption data, which is able to accurately identify the load operation status. Furthermore, the algorithm is shown to be highly efficient and can be realized via fast Fourier transform (FFT), and its hardware packaging can be easily realized. © 2017 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.